Research On Path Planning Algorithms For Unmanned Aerial Vehicle

With the rapid development of information technology,aviation technology has been developed rapidly.The complexity and variety of the flight environment pose great challenges to the maneuverability of the Unmanned Aerial Vehicle(UAV).Planning algorithm is the keystone of flight path planning,and the key technology of automatic flying.Therefore,the path planning algorithms for UAV is worth further developing in both theory and practical applications.This dissertation focuses on designing planning strategies.The main contributions of this dissertation are as follows:The maneuverability model of the UAV is established to meet the constraint conditions in the flight path planning.According to the operational requirements,the static threat sources,such as fixed radar,in the flight environment are analyzed and the corresponding models are developed.Meanwhile,a comprehensive cost function model is constructed,which is the foundation of path planning in the following parts.Further more,path planning in static environment is investigated based on improved Genetic Algorithm,in which a new encoding method and operation are designed.To be specific,a real number coding method based on the change of course is designed.The temporary path is formed according to the individual code vector.Then the simulation results are provided.Path planning in dynamic environment is further studied.In this context,a fusion algorithm is proposed based on Interactive Multiple Model(IMM)algorithm and Model Predictive Control(MPC)algorithm.For avoiding the moving obstacles,the IMM algorithm is used to predict the trajectory of the moving obstacles according to their current positions.Besides,Converted Measurement Kalman Filtering algorithm is used to converse nonlinear system into a linear system.Considering the existence of both static and dynamic threat constraints in the flight environment,the real-time performance and effectiveness of the proposed fusion algorithm are verified by simulation.Taking path planning in the three-dimensional space into account,an improved ant colony algorithm is established.Firstly,3D Flight environment is built by using digital map technology.Then,the integrated cost objective function is given under three constraint conditions: the threat cost,the distance cost,and the high cost.Besides,the state transition probability is reset based on the distance between the selected node and the target point.In addition,in terms of the update of pheromone concentration,the classification method is employed to set the rang of pheromone residue factor,which can improve the diversity of the search path and avoid the algorithm premature,decreace the searching time and avoid converging local optimal solution.The simulation results show that the obtained path based on the designed well meet the desired one.